The Presumption of Racism

Antony Flew, a celebrated 20th century philosopher, put forth an argument in favor of atheism which states, roughly, that in the absence of any clear evidence for theism, atheism is the rational presumption. I would quibble with this argument, for such a presumption is inherently tied to a positivist view and simply fails to take the idea of God seriously [Flew, it is worth noting, later became a theist after engaging with Aristotle]. However, this model of presumption that lacking evidence for some specific thesis is evidence enough to establish the rationality of disbelief is not necessarily flawed with respect to theses of a less metaphysical scope. For instance, a presumption of “There are no invisible pink unicorns on the dark side of the moon” is something we would consider reasonable, and it would require extremely different metaphysical priors for a nearly-certain probability to not be assigned to that belief. We might consider instances of failure to appropriately weigh negatives to highly specific, highly unlikely instances indicative of a general failure of rationality. That is, such a failure at the fringe might be considered akin to debugging; if you’re unable to both state what would constitute evidence of such a unicorn and to then elaborate on why it would be so unlikely to find evidence of such, you are unlikely to have a solid grasp on sound principles of methodological analysis.

On the other hand, if we were to find such a unicorn on the dark side of the moon, this would highlight that our ordinarily functional rationality has a large blind spot. But, assuming there is no such blind spot [in this specific area of our rational periphery], it proves itself a useful kind of test.

Granted, it is likely almost no one wouldn’t be willing to weight the negative thesis with near-certain probability; it illustrates a principle at an extreme degree. We may also apply such forms of “rational debugging” on other people in more specific ways. It is very easy to have a view of the world which allows one to conclude the near-certain improbability of unicorns being on the dark side of the moon; it is much more difficult to elucidate complex social systems. Consider the following description, which will function as a kind of theoretical litmus test in which I have removed certain key elements.

A specific subpopulation of the greater society is disproportionately represented in prisons.

The explanation you’re likely to reach for is likely to reveal your ideological priors. There are two positive theses which present themselves as likely explanations: either this represents something innate of that population, or it demonstrates that the greater society has a bias towards putting members of that subpopulation in prison. Both of these explanations have concrete instances we can conceive which would serve to verify either of these hypotheses. Those who would tend to offer either explanation and then stick to it despite further information are revealed to have a defect in their rationality; in reality, the above description isn’t sufficient to verify either explanation. The proper response is to admit of both possibilities with an agnosticism dependent on further information, if forthcoming [with a caveat that this might still warrant avoiding that subpopulation if you are unable to access further information].

If the specific subpopulation in question were men, who are indeed disproportionately represented in prison, we are likely to accept the former explanation. That there are more men in prison compared to women is explained by their being innately more aggressive and having a lower aversion to risk, traits which lend themselves to a higher likelihood of criminality. At the same time, we also notice that men are likely to receive harsher sentences than women for the same crimes, which we attribute to this innateness; the more likely a specific subpopulation is to exhibit criminality, the harsher the punishments must be in order to act as a stronger deterrent of that behavior.

Now consider if the specific subpopulation we were speaking of are blacks. Whereas the explanation for why men as a group are more likely to be in prison than women is explained by innate differences, the suggestion that disproportionate representation by blacks is due to innate differences is racist. There is, in other words, a kind of presumption in favor of racism we exercise which we don’t exercise when it comes to men. This isn’t necessarily an indicator of a failure of rationality. After all, we do recognize cases where certain populations are prejudicially incarcerated and overall persecuted. The Nazis persecuted the Jews, Rome persecuted Christians, the Catholic Church persecuted the occasional atheist, Christians persecuted Muslims, Muslims persecuted Christians, so on and so forth for almost every potential pairing of groups. If we find a disproportionate representation of Christians in prison in a self-proclaimed secularistic society, the only innate factor we might attribute to the Christians in this case is a religious zealotry, but not innate factors that make them otherwise predisposed to criminality under sane conditions of legal jurisdiction.

However, our means of distinguishing between which hypothesis serves to explain the disproportionate representation of a subpopulation in prison requires the elucidation of a theory which allows us to test for the mechanisms in place which verify the former or latter hypothesis. If one takes the mere fact of disproportionate representation as proof of unjustified discrimination [disproportionate representation by a group can also occur due to justified discrimination, e.g. those employed by NASA are disproportionately represented by those of 125+ IQ, and with good reason], then we must be forced to conclude that the disproportionate representation by men in prison is not due to innate traits such as greater mean aggression and lower mean risk-aversion, but is the result of a mere prejudice.

After all, the feminists do tell us that women are psychologically essentially the same as men, and criminality is a psychological trait. Either men are overly represented and/or women are underrepresented. Indeed, there are some feminists who even make the argument that patriarchy is a net disadvantage to men, with only a minority of men being benefited; this could be evidence of just such a patriarchy. I don’t know how this male-female disparity could be solved, but it’s worth pointing out that this disparity stands in need of a rigorous, scientific justification.

That said, many would still, and quite reasonably, remain with the hypothesis that men just are more innately prone to criminal behavior. Vis-à-vis criminality, men have a higher mean than women, which is explained by simple biology. Once you accept such an explanation, however, this same theory could also explain the disproportionate representation of blacks in prison. At least in principle, if the theory is sound in terms of explaining the gender disparity, it may explain other group disparities such as racial disparity.

If we are to distinguish between biological and social reasons for group disparities, then we must have some way of measuring those differences. Biological differences in psychology appear to be measurable; for example, exhibited time preference in young children has a correlation to life outcomes. If we are willing to accept that this experiment documents a real, genetically influenced trait, this presents a very simple means of finding evidence for weighing the difference between biological and social means of explaining racial disparities [it must at least document something strongly ingrained by the age of the children being observed]. If no significant difference is found between children by race, this is evidence in favor of the social hypothesis; on the other hand, if there are significant disparities found between race, and those disparities also happens to approximate the differences we find elsewhere in society, that is evidence in favor of the biological hypothesis. [I’m aware of potentially confounding factors; for instance, there was a high correlation between socioeconomic status and ability to delay gratification. This might potentially be explained by the conditions of poverty socializing a child into lower ability to delay gratification. However, this can be controlled for by simply comparing children of similar socioeconomic backgrounds. If anyone can find a study which does so, please send it my way; I searched my usual academic sources and could find nothing.]

Back to Flew and his presumption of atheism. Borrowing a parable from John Wisdom, he writes:

Once upon a time two explorers came upon a clearing in the jungle. In the clearing were growing many flowers and many weeds. One explorer says, “some gardener must tend this plot.” The other disagrees, “There is no gardener.” So they pitch their tents and set a watch. No gardener is ever seen. “But perhaps he is an invisible gardener.” So they, set up a barbed-wire fence. They electrify it. They patrol with bloodhounds. (For they remember how H.G. Wells’s The Invisible Man could be both smelt and touched though he could not he seen.) But no shrieks ever suggest that some intruder has received a shock. No movements of the wire ever betray an invisible climber. The bloodhounds never give cry. Yet still the Believer is not convinced. “But there is a gardener, invisible, intangible, insensible to electric shocks, a gardener who has no scent and makes no sound, a gardener who comes secretly to look after the garden which he loves.” At last the Sceptic despairs, “But what remains of your original assertion? Just how does what you call an invisible, intangible, eternally elusive gardener differ from an imaginary gardener or even from no gardener at all?”

 

Growing to Fail

There are 2 conditions required for games to reach positive evolutionarily stable equilibrium:

  1. Personal and collective interest are aligned with each other
  2. Cooperation has positive-sum gains for players

When speaking of society, which might be described as a conjunction of different games all operating simultaneously, we have generally struck upon positive evolutionarily stable equilibria; had we not, we wouldn’t be here in the first place. The more people in a society tend to cooperate, the greater the collective gain. This is why the tendencies of individuals, with even a slight shift of their game theoretic strategies, can do so much to explain the differences between the prosperity of societies. Why is the United States so much wealthier than Liberia? At least, perhaps the greatest part, of the explanation has to do with how likely individuals are to cooperate. In a society such as Liberia, there is simply more defection taking place, which wears at the possibility of an industrial ascendancy. Then again, within a society such as Liberia and given one’s own personal assets (e.g. little to nothing), defect strategies are the most rational strategy for at least a great deal of individual players. (Both conservatives and liberals would do well to remember this point. When everyone’s defecting, your personally optimal strategy is defection.)

Not all positive evolutionarily stable equilibria are as positive as others. On the other hand, a society such as Liberia, which is more effectively individualized (i.e. individuals are more able to survive on their own without having to depend on others’ cooperation), is also much more resilient to global, or external, shocks. The right confluence of disasters, natural, economic, and social could cause a larger decline, or even collapse, of US society. Liberia, on the other hand, is relatively resilient. Sure, Liberia has less to lose in an absolute sense, and undoubtedly a collapse of global networks would wreak some havoc, but it is more likely that the equilibrium which holds at present in Liberia will re-emerge more quickly than the equilibrium we have in the US.

Humans, as social creatures, are dependent on integration with a group for reproductive success, and in probably most cases, for survival. We are evolved to be ready to cooperate and to expect cooperation from others who have sufficiently signaled their parallel intentions. A stabler society in which defection is less frequent is a society in which it makes more sense to individually invest in highly specific forms of labor which, outside the ability to trade that labor for highly liquid assets (i.e. cash), no such investment would ever occur. At the same time, this makes that individual even more dependent on the cooperation of others for achieving his goals. However, we seem to have struck a balance that has allowed such a proliferation of specialization, and many have brought themselves out of poverty through this macroscale effort at cooperation.

The coordination of so many individuals spontaneously is not, and can never be, overseen by any intelligence short of omniscience. Society is not the product of intelligent design, but just like our own bodies, is the result of natural selection. What is being selected for? Spontaneous means of playing the games that constitute society in a way which is most likely to produce positive evolutionarily stable equilibria. Those forms of society which are, by comparison, relatively disordered shall inevitably be selected out, some quicker than others. Given the dependence of humans on each other for survival and reproduction, a large decline or even a collapse of society would not entail the end of humanity, only a meaner existence nearer the soil.

There is a problem to positive-sum games. Provided an evolutionarily stable equilibrium takes hold, such a society shall tend to grow until it is confronted with a higher-order game that its forms of coordination (e.g. social norms) are not adapted to. With growth comes increasing complexity, increasing complexity entails more sophisticated games which are, in many ways, beyond the ability of individuals to perceive, let alone begin to solve. And even were a solution to be known (“everyone should cooperate by doing x!”) it is generally impossible to arrange for everyone to spontaneously begin cooperating in this way except through a centralized fiat, and even then there are strong limits to what that can achieve (the anarchist in me wishes to note that the centralization of a particular social response may eventually prove evolutionarily outmoded, e.g. the future might be stateless). The initial growth of the society secured by its maintaining the requisite social norms inevitably places it in a confrontation with a collective action problem those initial social norms are not adapted to, and nor are these social norms capable of instilling a new coordination which will prove a positive evolutionarily stable equilibrium in response. The doom of a civilization is written into the social norms that first allow it to rise.

At some point one or both of the conditions necessary for positive evolutionarily stable equilibria are unable to hold. Personal and collective interests diverge, or cooperation ceases to have positive-sum gains for both players. There is nothing in the world that guarantees both of these conditions shall hold indefinitely. If a new technology to increase economic productivity in proportion to population growth isn’t forthcoming, then at the very least the potential for materially positive-sum games is vastly diminished (there may be localized deviations from a net zero- or negative-sum of games). New technologies may even disrupt the regeneration of social norms in new generations, if children’s and parents’ interests diverge, which brings about more divergences (especially in intersexual relations; we can’t forget, nor is it useful to pathologize, that men and women have innately competing personal interests, i.e. polygyny vs hypergamy). The tendency towards lowering time preferences which is otherwise exhibited by long-lived institutions or series of generations will ironically frustrate the attempt to do so when a fitness valley is reached and the proper response isn’t available to the group.

Game Theory, Carcinisation, and the Omega Point

Game Theory, Carcinisation, and the Omega Point

Game theory: study of strategic decision making.

Carcinisation: a hypothesised process whereby a crustacean evolves into a crab-like form from a non-crab-like form.

Omega Point: the purported maximum level of complexity and consciousness towards which some believe the universe is evolving. The term was coined by the French Jesuit Pierre Teilhard de Chardin.

Humans are a very strange creature compared to the rest of the animal kingdom. The primary form of competition and selection is imposed on an individual by his own, or another’s, society. While the occasional extraordinary natural disaster in the form of a tornado, earthquake, viral epidemic, and so on wreak their havoc on humans and human societies, for the most part we have mastered the forces of Nature, at least insofar as we consider those forces external to human society. The primary source of evolutionary selection for the human race, since the dawn of civilization (perhaps even before that; maybe even at the Dawn of Man [if you’ve never seen 2001: A Space Odyssey, well… take some acid and watch that asap]) has been imposed by humans themselves on each other. You are less likely to be impeded in your success at sexual reproduction by merely natural forces than by the social systems which evaluate your sexual and moral worth as a human being.

This suggests that the limits of our growth as a society is imposed less by mere forces of physics and biology, which we have found ways of working around or even incorporating to our advantage (not to discount that they remain hard limits), but by the problems of coordinating many rational actors. Not all spontaneous actions undertaken by individuals, whether they impose a burden to the individual or propose a benefit, are aligned with the collective’s survival. At times, the individual’s personal interests and collective interests are perfectly aligned, and we do not need to augment their implicit game theoretic strategy through incentive structures, but at other times we are only able to develop stable Nash equilibria for the collective action problems facing societies through elaborate traditions which punish defectors and punish those who won’t punish the defectors in ways that are difficult to detect for the uninitiated. Over time, as we gain more and more mastery of physical and biological nature, we shall brush up all the more fiercely and intensely with the limits of the collective action solutions we have yet been able to maintain. Many times, there is a cycle of integration and disintegration, wedded to the dynamics of population, conflict, and production, which sees simpler collective action problems find solutions, but this bumps the problem back another step and introduces the nascent society to higher-order collective action problems it has never been confronted with. This is a fancy way of saying that, once one problem becomes solved, the growth of social complexity eventually outstrips the ability of the initial solutions to cope, and society declines.

From a far enough perspective, however, these oscillations between social integration and disintegration yet suggest an upward arc. As Teilhard declares, “All that rises must converge.” While this seems an apparently Progressive kind of thing to say, it actually forms an inversion of the Progressive eschatology. Given Teilhard’s Catholicism, this is perhaps unsurprising. Is he right?

At the end of the day, there is almost always only one optimal solution. While some solutions to a given problem may be adequate for its time and place, a superior solution will eventually outcompete and displace that initial solution. At times, that initial solution is a necessary step in the production of the later optimal solution. This the dialectic of evolution, which whittles forms down to those best able to propagate themselves over time. Given the reality that along even a very high n-dimensional axis there shall always be a tendency towards equilibrium, this suggests that over time social beings tend to one particular form. In other words, given an infinite amount of time and assuming a continuous chain of being, there can only emerge one inevitable form.

Within limited and specific conditions, game theory predicts that particular games have one inevitably optimal solution. Assuming players of a game form a group, over time those groups which achieve greater game theoretic evolutionarily stable equilibria shall remain. For instance, the iterated prisoner’s dilemma optimal solution is the strategy of tit-for-tat; initially cooperate, but respond to defection with your own defection. Over time, and assuming a competent player, cooperation should become the norm. This cooperation in real life situations that approximately model this dilemma expands the territory, and in expanding the territory introduces new means of interacting and playing the game, producing the need for a more complex strategy. However, in all cases there shall tend to be a convergence; each game, no matter how complex, has one optimal strategy. (In the case of different strategies which, for that particular game optimize to equally maximal outcomes, keep in mind that in society the solving of one game introduces another game which needs to be solved; these consequent game chains are what will select for those that produce optimization at higher orders.)

All of this is to say that, ultimately, the universe is set up so as to select for some particular optimal game theoretic strategy employed by societies, which in turn makes it inevitable that there is one best form of society which will inevitably outcompete and displace all others. This “final” society, or conglomeration of self-propagating forms found at the end of time, is the hypothetical Omega Point. Whether one would like to call it equilibrium or teleology, it is suggestive of something (you don’t need to buy into Teilhard’s speculations that this Omega Point is God, but he is certainly right to point out this maximal level of complexity/consciousness is particular). From the perspective of God, whatever the universe is, given interactive social individuals have only one way to proceed forward (or else be selected out), the universe is an Omega Point making machine. From a certain perspective, the end of time operates to draw the universe to it as much as the beginning of time pushes the universe.

Malthus isn’t Dead: Population, Productivity, and Decline

Malthus, writing in the late 18th and early 19th century, struck upon a salient, elegant, and horrific socioeconomic theory. Assuming a fixed level of productivity, growth in population would drive down wages (more labor supply = lower wages) and increase rent (more housing demand = higher prices). Inevitably, this would result in widespread poverty and misery, since short of just killing a lot of people, the high demand for basic necessities will make them expensive relative the wage the average individual is able to draw. Granted, if you are one who owns the means of production, low wages and high rent is a formula for great and fabulous wealth (and so these periods tend to bring out the greatest wealth inequalities); but for the rest, short of an increase in economic productivity, the pie isn’t getting any bigger and the slices keep getting smaller.

This lends itself to a “Malthusian ceiling,” conditions in which fertility and death are at par, giving a society experiencing neither population growth nor population decline. Usually, population does not remain static; about the point that the Malthusian ceiling is reached, elite overproduction leads to too many chiefs, not enough Indians, and political instability leads to lower economic productivity which leads to population decline. Generally, the political stability that made it possible for a society to reach its maximum economic productivity eventually leads it to stagnation and decline. Rarely does a society just stagnate, however; competition between elites, in the case that the territory is no longer expanding, tends toward the greater employment of hawk strategies, which are zero- and frequently negative-sum. Then again, given there was no positive-sum game available to be played, it is only rational (from the perspective of an individual elite) to employ this hawk strategy. Enough of these hawk strategies, and eventually your social cohesion corrodes and the overall level of economic productivity falls with it, reducing many elites to poverty and many of those who were in poverty to starvation.

The decades following the moment when a society reaches its Malthusian ceiling tend to be Interesting Times, with the splintering of political factions, changes of government, and mass population movements. These oscillations between population growth and decline is what Peter Turchin terms a secular cycle. These secular cycles are tied to population and productivity, with relevant influence by external forces (for instance, the wholesale invasion and subjugation by an external force will abort a secular cycle). The upswing is the integrative phase and the downswing the disintegrative phase. Disintegrative phases do not always precipitate complete social dissolution, but typically people will be poorer, crime and war will be more likely, and generally it just isn’t as much fun.

We can grant that Malthus certainly did not see the Industrial Revolution coming, but we mustn’t let that obscure the truth he uncovered. The Industrial Revolution brought a historically unprecedented increase in economic productivity, but it did not suddenly bring us to a point of post-scarcity where population size has no influence on median wealth, inequality, or social stability. The overall increase of economic productivity was large, but at the end of the day it was only finite, and unless productivity growth continued at such a rate, population growth would inevitably catch up. Yes, we moved the Malthusian ceiling much higher, but it’s still there as much as ever.

Has productivity growth kept up with population growth over the last few decades? I’ll admit that I haven’t figured out the best way to measure the relationship of productivity and population (GDP? energy use? further analysis will take place), but it is certainly true that not all increases in population have equal marginal product. Sometimes, an additional mouth to feed comes without any additional increase in economic productivity. There is a point at which marginal product to population increase is negative. Outside growth to productivity, population will inevitably eat up your gains and reduce everyone to the same levels of poverty as they were in before that initial increase in productivity.

Where are we in the secular cycle, as influenced by the rapid industrialization of the last two centuries? We appear to be reaching peak population growth at the same time as our primary energy sources are becoming more difficult to access. The Millennial generation expects to be poorer than their parents, and it would be unsurprising if their children could expect to be even poorer. Wealth inequality has reached dazzling heights (note, for the uninitiated; I see nothing wrong with inequality per se, nor is inequality anything but epiphenomenal to the underlying causes) and it doesn’t appear that it will fall any time soon. There is an overabundance of elites and aspirational elites, and their means of whittling each other down is becoming increasingly hawkish. In short, based on back of the envelope calculations comparing the above described framework to present conditions, we are about to enter a disintegrative phase of the secular cycle.

When Doves Become Hawks

The hawk-dove game is a very useful dynamic for illustrating the conditions under which cooperation between individuals will thrive, and likewise those conditions under which hawks will thrive. In this game, whenever a resource may be potentially contested, players may choose a dove or a hawk strategy; the dove strategy entails cooperating, sharing a resource between the two, while the hawk strategy involves defecting, intending to fight off the other player to keep all the resources to oneself. If both players choose a dove strategy, the resource will be shared and no cost in the form of violence and effort exerted takes place. If one player chooses a hawk strategy, and the other a dove, then the hawk will automatically get all the resources. If both players choose a hawk strategy, then both players will have to fight, with the winner getting all the resources and the loser having nothing.

As a society, we want people to all be doves; beyond the hawk-dove game itself, hawk-dove dynamics go a long way to explaining the potential outcomes of contested resources (and don’t think about this only in terms of material resources). Many times, the benefits to cooperation, i.e. both being doves, is positive-sum. When a sufficient number of the games one plays in society are positive-sum, then doves proliferate, as over time the doves cooperating with each other gives them the resources to shut hawks out of the game and bringing about a second-order equilibrium. So long as expansion is possible, whether an expansion of territory, an expansion of resources, an expansion of opportunities, whatever, doves in the long run will outcompete hawks provided they also hold to a “punish the defectors” strategy. (Criminal law is an example of a “punish the defectors” strategy. But there are many real world examples besides this.)

However, given a continuing expansion of resources available to all players is the condition under which doves proliferate, what about conditions in which no such expansion is possible through cooperation? Granted, even at a zero-sum equilibrium doves can rule so long as they continue to successfully employ a “punish the defectors” strategy, but maintaining this strategy will become dependent upon a third-order strategy of “punish those who won’t punish the defectors.” Why? Because the doves are not in a position to outcompete hawks merely through cooperation, given there is a fixed amount of resources. Maintaining the first-order equilibrium of dove strategies requires that even where a hawk strategy is successfully employed (netting a positive gain for the hawk initially) punishment can be secured. The securing of punishment is dependent on the second-order equilibrium being maintained, and void the possibility of the doves simply always being wealthier and more powerful than the hawks, that second-order is dependent on the third-order equilibrium being maintained. If those who choose to punish the defectors (at a cost to themselves; punishment is not a costless endeavor) don’t see that everyone else is contributing to the overall punishment of defectors (secured by the third-order of punishing those who won’t punish), they may choose to no longer punish defectors. As soon as this occurs, hawks increasingly “get away with it,” and thus rapidly proliferate and displace the doves.

Some people will choose to employ a dove strategy until the end of time, even as it becomes increasingly obvious that doing so ensures their takeover by a hawk (compare: UK and Muslim culture). However, most people are of a kind that they will simply choose whatever strategy appears to be most likely to bring them gain. In a world where the third-order equilibrium of punishing those who won’t defect is maintained, these people will very easily make the choice to be doves. Likewise, there are some people who will always choose to employ a hawk strategy, no matter the likelihood of their punishment and negative return. These people are the reason the third-, rather than merely second-, order equilibrium is so crucial. Ideally, a third-order equilibrium provides positive gains to punishing the defectors (i.e. cost of punishment is shared, individual acts of punishment bring reward), and so hawks will find themselves most attracted to these positions, as it suits their abilities and proclivities. In the real world, these are the arenas of politics, war, law, and crime enforcement.

What precipitates the turning of doves into hawks is when the doves notice that hawks aren’t being punished. In a game with millions of players, this can take a while; the isolated case of collusion between chosen law enforcers and criminals may initially be written off as just that, an isolated case. Noticing when such collusion, or the mere lack of enforcement, becomes the rule rather than the exception, is difficult, and like many other macroscale phenomena of societal breakdown, isn’t noticed by the mass until all havoc has already broken out and it’s too late to do anything. But the complete breakdown of social order will not tend to occur until after the second- and third-order operators (where violence is exercised legitimately) cease cooperating with each other. So long as expansion, or at least the maintenance of order, is apparent, dove strategies will rule among even hawkishly inclined individuals.

Endless expansion is, however, impossible. There are cycles of society tied to technological increase in productivity and population which oscillates between easy growth and hard declines. An initial opening of expansion (e.g. suppose 1/3 of the population dies off, some new technology increase productivity by a magnitude) very quickly engenders dove politics. Upwards mobility takes place in the absolute (and frequently, also relative) sense, and elites have enough territory at their disposal to satisfy yearnings for power such that the interests of politicians and society are aligned. What must be kept in mind is that politicians do not fight through personal combat; inasmuch as war is politics by other means, politics is war by other means. Elites who intend to maintain positions of elite status while the territory of power is no longer expanding even while just as many try to obtain these positions of elite status will be increasingly incentivized to employ hawk strategies against each other. There is, after all, simply less territory available to each, and so territory gained by one frequently comes at the expense of another.

How do elites fight? Given they have at their disposal social capital to employ, the means of fighting with each other are necessarily through social means. Fighting, we must remember, is never costless; something must be given up in order to fight. With society itself as the very battleground, this will frequently be to the advantage of those willing to destroy more. Were an initial transgression to be responded to in kind, somebody will lose something. The elites do not tend to bear the costs of their actions in a very personal way, as most of their costs are externalized. This means their fighting against each other is a cost imposed on society.

If material conditions changed, such that there is simply no more territory to expand into, doves will not be able to maintain dominance through their cooperation. Hawks increase in number, and with that, given the fighting of elites entails the expenditure of social resources, overall sociopolitical stability is lower. Until expansion is possible again, hawk strategies proliferate, sociopolitical instability becomes the rule, defection entails cascades of more defection, so on until the hawks have eliminated each other and a group of elite doves is able to orient society towards expansion again.

#GamerGate is an Ideological Litmus Test

I didn’t really want to write about GamerGate, and in fact I’m still not going to; you’ll have to forgive me for being a little clickbait-y. In fact, I have preserved a virtually complete ignorance of the details, and am proud of that. So far as I care, it is a personal spat, no more significant than any other fight that has occurred between former lovers. I already know the feminists and their bastions in the mainstream media will circle wagons around Zoe Quinn, taking advantage of the apparent narrative of oppression to beat the bad guys up with it, while the gamers who pay attention to video game media will have complicated, nuanced views like you’d find on Reddit [that’s not a compliment, by the way] attempting to mount a defense of their freedom to associate with people even if that doesn’t happen to select for gender equality in a medium’s intended audience. It has gained its prominence not because what occurred was so vastly shocking [if shocking were the only factor in memetic spread, the internet would be full of nothing but gore and porn with the random Holocaust denial thrown in for variety], since even the most precursory perusal of a city newspaper will reveal transgressions that amount to more than emotional attack, i.e. actual physical harm and destruction. No, it has memetic success because it is perfectly designed for propagation through instilled narratives of oppression, a kind of memetic Greek fire, if you will, for serving as an exosemantic gang sign. It provides the same function as the wearing of handkerchiefs in signaling association with a gang.

The internet, and by extension the communication achieved by humans through it, is of an informational kind. We are dependent on the survival of the groups we align with for the propagation of our genes, ultimately. Humans detect group association because group association is personal identity. This is especially the case on the internet as the entire content of an individual’s persona as transmitted through mostly textual format is essentially the opinions and ideas that person has expressed. On the internet, you are your opinion [and opinions are, ultimately, attempts at group association]. This, combined with the human instinct to figure out which side you think is best and thus to best identify with and promote the survival of your particular opinion group means opinions are akin to fashion. [And no, you are not above trying to be fashionable.]

This is the basis of memes, which propagate via the utility afforded to individuals by their use. That utility can be mere entertainment, intellectual edification, the promotion of effective life strategies, and the ability to identify which group of people you want to associate with, who you expect to treat you the best. Some opinions are more exosemantically loaded than others; for instance, one’s opinion on ice cream flavors signals relatively nothing about group identification, but one’s opinion on Zoe Quinn allow a person to simultaneously 1) identify with a group, 2) attempt to make a personal contribution, and 3) receive praise from others for the soundness and educated-ness of his opinion. As a meme, GamerGate is basically an ideological litmus test; are you a bad enough dude to declare yourself a feminist/anti-feminist? [Nor am I going to act like I’m superior for being “above it all”; reflexively, I assume that 4chan [RIP], manospherians, and gamers who don’t care to signal moral superiority have the best grasp on the details and how facts have been selected, omitted, exaggerated, and emphasized so as to fit a particular, and well-known, narrative about an oppressed damsel in distress who was just trying to be a strong, independent, grrrl. I might as well point out that you already had your mind made up about whether you agree with me or not before you even read this; you’re only reading at this point in order to either continue uncovering my awfulness as a human being and confirming your own moral superiority or to find reassurance that you can be a gamer without having to make gaming about feminism. Among other things, pointing out a woman’s infidelity in previous committed relationships is a public service to men who might otherwise be duped into giving commitment when they might not if that history were known. As to other tactics, harassment, humiliation, again I am ignorant of the particulars, and have no care to know.]

Zoe Quinn, and her boyfriends or cock carousel comrades, all of this is boring. Sans GamerGate, Zoe Quinn would have remained forever a nobody, a woman who tried to leverage her sexuality for material gain and got caught. The news has ceased to be about Quinn, but is really about itself; it is about the implicit need for not yet politicized arenas of social gathering to be pressured to conform to ideological expectations. The detection of insufficiently leftist memes propagating in social groups is the news here, and while the insufficient leftism of gamers [who are predominantly male, which entails the automatic suspicion of sexism] has been well-known and discussed since the introduction of graphical power sufficient to realistically approximate female breasts. It simply happens to be the case that there hasn’t been a sufficiently awful casus belli to use as an excuse to excoriate, shame, and I’ll just go out on a limb and assume eventually attempt to fire game developers, journalists, well-known gamers [which could have already occurred for all I know about this particular case] as punishment for insufficient leftism of the video game scene. GamerGate isn’t about Quinn in the least; she is merely a convenient springboard for diatribes on the awfulness of patriarchal oppression and the shaming of those who don’t conform their own opinion with sufficient feminist orthodoxy.

It was the same with Ferguson. White cop shoots unarmed black youth, the narrative of racist oppression is obvious enough that it spontaneously coalesced opinions and media was on the story like sharks detecting blood. Trayvon Martin is another character perfectly formed for ringing the bells and declaring a panic about racial relations. Rodney King was, with respect to the 24 hour news cycle that hadn’t yet developed and an incipient internet that was only beginning to filter down to the masses hadn’t yet formed stable opinion groups the average individual could identify with. The Lewinsky scandal is another primitive form of ideological litmus test; just look at the Republican/Democrat split of opinion.

There’s another case from a few years ago which is an illustration of the exact same tactics to attack a group that wasn’t expressing sufficient leftism. Anyone remember ElevatorGate? Long story short, woman experiences oppression by man [once again, I largely avoided knowing the details of this case, but I still ended up learning a few things through osmosis], which sparks the question of whether the New Atheist community associated with Watson is sufficiently leftist; the answer, as you should know by now, is that no, not in the least. [The question only gets asked by leftism if it gets to prescribe its favorite solution: apply more leftism!] One’s opinion on ElevatorGate was a shibboleth to prove one’s requisite leftism in certain communities.

These outbreaks of panic about groups which fail to express sufficient devotion and tribute to leftism are largely epiphenomenal, the result of more fundamental sociobiological, cultural forces. Civilization never truly quelled the war of all against all which is the animal kingdom, it only negotiated a more pro-social, less violent form of conflict resolution, e.g. law, morality, social norms, innate bias to desire social belonging. These campaigns to uncover and root out individuals who express opinions which deviate from most others’ is a time-honored tradition in humans, with varying forms of disconnection from society entailed as a consequence; in some cases it is losing and/or not being permitted to gain a high status career, ostracism and exclusion from the community and family, or even death. [If you think this is intrinsically wrong, then you should oppose leftist calls for people who say the wrong things about gays or women to be fired from their jobs.]

What’s the point of this? There is a distributed conspiracy enforced by a media backed by academia [consider that you generally need a Ph.D. to be considered A Person Whose Opinion Matters in the context of social scientific claims; compare to the clergy of the Catholic Church in the Middle Ages] which stringently polices itself, and others, to only say things which are compatible with the reigning orthodoxy of the elites, i.e. Progress. Call this distributed conspiracy what you will, but it is a corroborated thesis that leftism, understood as a memeplex, propagates itself by the mechanism of concerted attack on groups and individuals that demonstrate insufficient leftism [see above, and consider your own reluctance to air your honestly believed but insufficiently leftist opinions for fear of attack]; it seeks out witches to shame [consider that finely honed sense you have for opinions which are borderline racist], attacks until submission is obtained, and makes sure those who aren’t persuaded stay shut up, ensuring the appearance of uniform agreement. This Zoe Quinn case is just an example of the leftist memeplex at work, and that is why boring details about a failed relationship has become Important News You Need To Know About. This isn’t even news anymore, but the simulacrum thereof; the news is constituted completely by the news, and this was but a memetic Kardashian in that context of meta-news media. These media scandals are just a way of proving you know how to say shibboleth, and simultaneously serves as a distraction from issues that might bring demotist pressure on leftists.

However, it seems the Cathedral overreached in this particular case, and the narrative of journalistic nonpartisanship is fracturing; journalists are being revealed for the zealots they are, people who place agenda above facts. The discussions which take place out of sight are the more important for revealing actual intention, and quite revealing they are when brought to light. That said, it should be kept in mind that any complicated network of relationships which successfully forms a feedback loop for propelling some particular ideological agenda through, say, media, are spontaneous orders. They develop not on the basis of some initial conspiracy, but simple interpersonal relationships, individual agendas, personal interests, which happen to develop into resilient social orders capable of achieving collective interests. In other words, insofar as any collective interest depends on the alignment of that collective interest with personal interests, those collective interests will tend to form. This is what a distributed conspiracy looks like, and it is indubitable that many other such collective agendas are rampant in media, with similar networks of connections to other industries about which much ink is spilled.

The Face of Gnon

The Still Face experiment is one of the greatest psychology experiments to have ever been performed, and through its well-grounded replication* it is one of the sturdiest foundations for theory in any form of the social sciences, and a theory of social psychology that could not incorporate this observation would almost certainly fail to make any accurate predictions about macroscale society. Time and again in my speculating and theorizing on social feedback loops I come back to the Still Face experiment as an analogy for the patterns and regularities of individual and institutional behavior.

*The experiment is very easy to verify. Simply try it on any available infant. And yes, I understand that it appears suspiciously abusive. I’m not necessarily condoning the experiment.

The infant, receiving the expected feedback, continues to remain happy and show behaviors associated with positive regard to the mother; but, when the mother ceases to express any facial expressions whatsoever, the infant becomes wary without the expected feedback. As suggested by Tronick, one of the original academics to perform the study in the 70’s, “an adequate explanation of the still-face effect had to consider not only immediate experience but also the infants’ appreciation that the expected patterns of social interaction had been violated.”¹ The infant, dissatisfied, indeed truly distressed by this state of affairs, takes action by all learned means he knows to obtain the desired result. He amplifies his signaling slowly at first, but finding that still produces no feedback, he continues to amplify until he has exhausted all potential means of signaling to find the expected, and strongly desired, feedback; in this case, the affection and approval of his mother, to know that he is recognized and cared for.

It is not difficult to extrapolate this model of behavioral amplification to other everyday social interactions between adults. The difference is that adults have had more time to learn and adapt a plethora of social behaviors, and thus understand how to most appropriately amplify the power of one’s signal in everyday social contexts. This can make the effect of amplification less noticeable, as they have learned better how to modulate their behavior in order to amplify their signaling power which will produce the best social outcome; compare a teenaged girl threatening suicide to a man answering the phone with a “Hello? Hello..?” The former is clearly more dramatic, and  thus represents a riskier escalation of signaling, as it may result in unwanted attention in the form of concerned and disciplining parents; teenage girls frequently threaten suicide in order to get attention, but not necessarily everyone’s attention. The former is no less a form of amplification of signaling power, but is many magnitudes lower in degree of strongest signaling strategies learned. The first hello, followed by the second with an inflection is a response to a lack of expected feedback; typically when you receive a call, you expect someone to respond after the first hello. Amplification of signaling power is, as you can see, of a morally orthogonal character; ”Excuse me’’ said as a means of getting a person’s attention is polite, while punching them in the face is not. The virtue of the behavior in question aside, the point is that this tendency to amplify signaling power in the unexpected cessation of feedback is an adaptive behavior.

This reflexive repetition is, as any heuristical behavior would be, not necessarily perfect, only good enough. Sometimes this repetition with amplification of the signal has significant and personally negative consequences; after all, some people really do kill themselves as part of a bid for attention.

In both interpersonal and institutional interactions, the employment of an amplification strategy to produce an expected and desired result will have either of three results; 1) the expected result obtains, 2) one gives up continuing to employ that behavioral strategy, or 3) negative consequences occur which forces the individual to stop its behavior. Sometimes the expected feedback just needed a little more push, and thus the repetition did happen to be truly instrumental in having a causal effect; sometimes the lack of expected result precipitates quitting or changing strategy; and sometimes they influence events through their signaling so as to bring harm on themselves. “Fighting words,’’ as it were. The strategy of amplification can be seen in the taunts leading up to a violent confrontation, countries threatening war, websites raiding and/or trolling each other, or corporations amassing patents, each being a signal of high strength and a strong claim of status, backed with the threat of physical, nuclear, digital, or legal means of resolving conflicts.

Forms of amplification which spontaneously order themselves and involve not strictly demarcated groups of people, such as nations or businesses, or even coherent signals, but those associated as a cultural group or thede, come in the example of Ferguson, Occupy Wall Street, and Chik-Fil-A Appreciation Day. In all cases, the formula is simple; initial action does not produce the expected result, so amplification of that behavior follows. The blacks in Ferguson had the belief that amplification of their protest via looting and rioting would bring reward, and perhaps it will. Occupy Wall Street was essentially a purely cultural phenomena, embedded completely within and inseparable from the general activist culture, explaining at once its failure and just what it is; it had no clear signals, as the signals themselves were revered as idols capable of producing second-order effects in the shape of appealing to the Powers That Be. It is a culture which is tapped into that fundamental willingness of humans to adapt behaviors they do not perceive any essential effect on the world, but do so in order to gain the approval of the group. Given the dependence of the individual on the group for the propagation of his genes, it is obvious social-historical evolution would select for an appreciable rate of credulousness, exploiting that heuristic of amplification through appeals to powers over-ascribed agency; religions which seek agricultural bounty in their gods and appeal to them through sacrifice and submission to the social norms associated with pleasing the gods.

Over time, the evolution of societies selected for those religions to survive which promoted the survival of its adherents. This might be considered how the instinct for religiosity developed, and the association of that sense of deeper power than what is immediately seen with particular symbols which help identify fellow adherents; the tendency to participate in and enforce social norms that have no obvious first-order benefit will arise over time simply due to groups which don’t being out-competed in the long run.  However, the fate of the society is substantially determined by the initial social norms it adapts; to borrow from genetics, social norms are a society’s genotype, which interact with material conditions in the form of nature and competition with other cultures, groups, and religions. It is probably useless to speculate as to the exact figures, but it is clear that what is particular to humans is that the social conditions exerted by human forces have had more effect on sexual selection than nature likely since the advent of man. The propagation of your group simply by relying on nature is easy, but humans have filled their ecological niche to the point that outside growing the niche through continual mastery of nature the greatest pressure on shall mostly come from other individuals and groups.

The selection for conformity to the group so as to best ensure genetic propagation has proven so effective in humans that we have been able to put together societies composed of more than a billion relatively long-lived, resource intensive individuals. This has proven possible through the ability to transmit highly compressed information, allowing individuals to transfer knowledge about the environment and shape the behavior of other individuals on the object-level with enough reliability that we essentially don’t understand the complete inner workings of a society and all its essential parts and functions. We have abstract models which become too fuzzy to be used on the actual ground level of human individuals actually interacting with each other. This creates a double-edged sword which evolution might exploit; humans which remain reliably influenced by social pressures can help propagate their group through taking advantage of vaguely or completely unknown second- or nth-order effects, but any weaknesses in that meme relative nth-order effects it fails to adapt for will subject it to its own deterioration while the fabric of the society it once unified comes apart.

The nearer the memetic functioning of a society, i.e. that implicit aggregation of the behavioral orientiation of all the individual members, to Gnon, the more effectively shall those memeplexes be whittled away in order to more rapidly induce norms that partake of second- and third-order effects, which resulted in sophisticated societies of impressive size and perpetuity such as Egypt and Rome within a relatively short amount of time since the rise of civilization. At the same time, such a size and complexity can extend initially constructive memeplexes into nth-order cycles of history that it isn’t adapted to, e.g. secular cycles.

If a vision or purpose of a society’s memeplex comes to take prominence in the distribution of social and material capital in that society, and that particular vision is inordinately invested in a particular cosmology which is at odds with adaptive intergenerational feedback loops, that society comes face to face with Gnon. The desire for the impossible, set above all other things, will not initially reveal itself, for the full effects of slowly tearing at the social fabric may not reveal itself for generations; that is how important social norms are to the effective implementation of collective active solutions for game theoretic problems faced by society which have feedback cycles measured in generations of human lives. A project originally set about to settle some problem, applied from a flagrantly inaccurate model of the world, will fail to bring about its intended goal; indeed, in many ways no particular goal is ever intended by any concerted effort of a group, but occurs reflexively through a memeplex’s adaptedness to the vagueness of society’s [some vagueness is good, as it permits submission to optimal practices with higher-order benefits we don’t perceive] conditions.

Consider a memeplex an abstract set of behaviors; now consider that set of behaviors as a single, information-dense composite signal. Given the formula of amplifying a signal in the case that less powerful signals of that kind fail to obtain the expected result, this predicts that with the consolidation of a memeplex’s eschatological vision of society, and the more fanatically it is able to divert resources to bringing about its impossible eschaton, then the response to failure shall be to amplify. No other strategy is left for obtaining the desired result; and you can wail and wail to the face of Gnon, but Gnon’s face will always be still. The only means by which the memeplex shall be destroyed is through the discrediting of the memeplex with the destruction of its proponents; being the ur-theory of all theories, the conditions of its own falsifiability is usually thousands of inferential steps from the speculative power of any human being. This effectively guarantees that a society captured by a memeplex shall be ruled by that memeplex, for better or worse, until its dissolution into barbarism or destruction by a competing group.

¹Adamson, L. B., & Prick, J. E. (2003). The Still Face: A History of a Shared Experimental Paradigm. Page 463.

Oppression as Game Theoretic Solution to Collective Action Problems

Sometimes the only solution to a game is to not play, such as games which involve mutually assured destruction in the case of defection. Sometimes the solution is to play it many times, such as the prisoner’s dilemma where iteration affords one the strategy of tit-for-tat, which will tend to produce optimal outcomes for both players. In cases where defection takes the form of free riding, the power to punish free riders often tends to bring about positive equilibria.

Consider a game in which a group of players are initially given $10, but given the opportunity to put their money in a pot where the total will be doubled and redistributed among the players. The optimal equilibrium for the players is everyone to donate all their money to the pot, where after being doubled everyone will receive $20. However, you do not know beforehand whether everyone will choose to cooperate; if you give your money and others do not, you will have less overall while the defectors will have more. After iteration, equilibrium tends to evolve towards little or no money being contributed to the pot.

However, if you introduce the power for people to punish other players, at a cost to themselves, a positive, near-optimal equilibrium tends to evolve through iteration. If you can give $1 in order to deprive $3 of someone else, there is the possibility of punishing defectors; when you realize that there are enough other people in the group willing to punish others for defection, the individual’s optimal strategy becomes cooperation, since defection will reliably result in worse outcomes.

If one has these groups compete against each other, with those groups “surviving” who remain in the top four quintiles of group success, one would tend to observe that those groups which most reliably establish this norm of punishing defectors will stick around, while those that cannot converge on this equilibrium will be selected out. Evolution as a group game, if you will. One notices that at this level of iteration, second-order feedback loops become possible; for instance, if one notices that, at the level of choosing to punish defectors, there are a few who yet reliably defect from that form of cooperation, the optimal strategy becomes to punish those who defect from punishing the first-order defectors. With the first-order game strategy guaranteed by an optimal second-order equilibrium, these first-order equilibria prove surprisingly resilient against entropic forces. Likewise, the second-order strategy entails a third-order strategy (punishing those who won’t punish those who won’t punish first-order defectors) all the way up, with these nth-order strategies becoming more pertinent with repetition and iteration assuming the group selection process (and repopulation of groups selected out from the populations of those groups which remain) goes on. Over time, assuming creatures which evolved to play only this kind of game, the ability of keeping track who has demonstrated a willingness to punish first-, second-, and nth-order defectors all the way up becomes necessary, as it would pose the only limit to how effectively a group can maintain itself once equilibrium at the lower levels is reached.

The moral of the story here is that, sometimes in order to solve collective action problems, punishment is an indisputable, occasionally essential, means of incentivizing cooperation.

When one examines society, there are a few social norms which are adoptions of this strategy. For instance, punishment for not paying taxes is a solution for solving the free rider problems which government ostensibly solves. However, an important caveat must be drawn here. In the above example, which is abstracted down to the fewest possible variables for the purposes of illustration, it is quite easy for one to both rationally perceive the dilemma one faces and how it can be solved. This ability to solve the dilemma through knowledge should not be assumed to be analogous in the implementation of collective action solutions; I quite purposefully introduced the notion of evolution in the equilibrium of individual games faced by these groups in order to draw out the fact that under the conditions of a particular game, evolution will solve to that implementation which is within reach. The group equilibrium can be obtained above without any conscious or purposefully intended implementation of a tit-for-tat, reputational strategy. Likewise, the actual implementations of social norms for the purpose of solving collective action problems have evolved largely void of purposeful intent. The theoretical problem of, say, maintaining armies is distinct from the historical and materially observable processes which were crucial for the practical solution of maintaining armies. Establishing a government in order to maintain an army is theory; the processes which allowed this to actually take place, and what allowed for certain governmental and social forms to proliferate over time was just blind evolution at work.

If I may with an apparent digression: how do you build a pencil from scratch? A pencil is, of course, a relatively simple tool for us to understand, but grasping the complexity of the practical task of constructing a pencil is many magnitudes behind any theory which could be reasonably elaborated. The first thing that is required is the materials; wood, graphite, rubber, and aluminum. But, the readiness and availability of these materials for repurposing together as a pencil is far too costly to go about producing these materials only for the purpose of constructing a pencil. In order to get the wood it must be cut down, in order to get the graphite it must be processed from a number of distinct products, likewise with the rubber and aluminum. The original reason for each of these products being available was in order to suit a directly relevant need which could be satisfied by the first-order pursuit of these materials. When someone was harvesting lumber, or producing graphite, or whatever, they were doing it in most cases because there was some directly perceived profit, not because they had in mind a pencil. It was the conditions of the availability of these items that led to the construction of a pencil; thus, we may observe that the complexity of practice is, after accounting for the processes which reasonably allow one to have access to the materials necessary for the production of some particular construct and the resultant repurposing of these materials. The theory of a pencil is extremely simple compared to the actual practice; but once a tool is successfully constructed once, the design may be distributed relatively costlessly, and thus do pencils enter society and become one of other tools and materials which may be repurposed to other uses over time.

I would propose the theory and practice of political philosophy face a similar divergence. The theory of political philosophy, which has for the last 500 years centered on the “social contract’’ model of spontaneous order through a conceptual individual’s analysis of the collective action problem facing society, puts to obscurity the actual practice. Given the utter absence of any possibility of an actual social contract, basing models of practice after conceptual practices which are impossible in theory unsurprisingly leads to a divergence between understanding and what is experienced. Ignoring that, in the sociological sense at least, every government was formed on through the original action of a few would-be sovereigns and any collective action problems a government happened to provide for and thus to secure the survival of its host society was a repurposing of that group. No government was formed in order to make sure roads would be built. The original construction of roads was because there happened to be a coincidental alignment of first-order and second-order profit in doing so; the first-order being that of, for instance, collecting tolls, the second-order being that it benefits your society by enabling more geographically diverse forms of cooperation. This being seemingly obvious, it cannot be forgotten that what leads us to suppose roads are the function of government is a traditional form of argumentation. Originally, no group was centralized enough to afford to build roads in the first-order sense, but once the first-order profit was possible, this instilled the second-order benefits which entailed the survival of those groups which best utilized roads. (Any controversy over the necessity of government building roads I leave aside for now.)

The crucial feedback loops necessary to the prosperity of a society is dependent on the alignment of first-order and second-order collective action solutions. Punishing defectors offers a first-order equilibrium, but punishing those who won’t punish defectors offers a second-order, and stronger, equilibrium. However, the knowledge of punishing those who won’t punish defectors is not necessarily so obvious as the knowledge that defectors should be punished, for it entails a level of abstraction which separates the event of punishment from its result, and is thus that much more difficult to defend. This is where most people get tripped up by too readily identifying some practice or social norm as oppressive. We are very good at identifying first-order feedback loops, but not very good at identifying the second-order, meaning we tend to lose sight of the second-order benefits a practice has which on the first-order analysis has a negative game theoretic equilibrium.

Groping Elephants and Unifying Theories

The story of the blind men, each coming into contact with separate parts of the elephant, and coming to different conclusions as to what the creature in question is a fable best understood as epistemology. Of course, each is individually wrong, yet not without being misled in ways which are, from our vantage point of the theory of an elephant in mind, reasonable and not without merit. I have no doubt that were a trunk my only point of reference, I might mistake it for a snake, or the rough, sturdy legs for a tree. However, with the individual perspectives of the blind men taken as a whole, what emerges is that each is aware of a part of which they do not perceive the whole.

Sometimes this story is misused to suggest that others’ perspectives are equally valid even if they differ from our own, ergo peacelovetolerance or something like that. The true moral of the story is, of course, that competing perspectives are sometimes not incompatible, but are perceptions of an otherwise unperceived whole. This is the same process at work in science; collections of observations are collated, a discipline attempts to provide a theoretical framework for explaining these observations, and eventually with luck someone will come along who understands, or at least grasps, the fundamental postulates of what each discipline is grasping at and can provide a single, elegant unifying theory. Where before there were an abundance of explanations proportional to the abundance of observations, suddenly all the observations can be placed into an ideal perspective once the whole, or at least something more nearly resembling a whole, is brought to attention and shown.

Newton did exactly this with his theories of physics. Of course, before Newton all the available data was not only obtainable, it tended to be known by nearly everyone who happened to engage with everyday life. The motion of the planets and the moon, the relationship between angle and velocity for determining where an arrow shall fly, and the falling of an apple was well-known and theories abounded which worked well enough to allow for accurate prediction. Some of the theories may have been implicit, in the sense that a skilled archer is not performing calculus in his head when aiming his shot, and some were explicated as models which accurately predicted such astronomic phenomena as eclipses of the moon. What Newton introduced was a unifying theory; he developed calculus in order to describe the action of objects in motion, especially those affected by gravity. He further postulated that the motion of an apple falling to the ground and the revolution of the moon were actually the same kind of behavior, proven through the ability of his formulas to describe both phenomena. His few simple principles of physics were robust enough to allow virtually every material phenomena observable to our eyes to be described with sufficient rigor that their future action could be predicted. What Newton ultimately did was propose, where many others were grasping at parts and proposing distinct explanations to what they perceived, that what each was perceiving was but part of the whole. Astronomers and archers, among many others who would’ve been familiar with the motion of falling objects, were both acquainted with but a part and did not realize their individual perspectives made the most sense from the vantage point.

To biology, Darwin accomplished essentially the same effect. Where you had a fossil record which indicated that many creatures once dwelt on the earth that no longer do, variation in species which included greater and lesser levels of similarity, sexual generation, and the knowledge that children had a tendency to be like their parents, the theory of evolutionary descent unified all these observations while simultaneously pointing the study of these parts in the right direction. It is truly staggering the unity evolutionary theory brings to the biological sciences.

At present, there is a paradigm shift taking place in the social sciences. The standard social scientific model, which presumes psychological sameness between human beings, is in truly dire straits, and is being forced to produce epicycle upon epicycle to explain away apparent contradictions. However, there are a number of thinkers working at the frontiers of social science, obviously where the institutionalized and domesticated progthinkers do not dwell, who are beginning to put together a unified theory of human society. The introduction of mathematical models from ecology for describing the growth, and decline, of human populations, the application of economics and game theory for explaining what social phenomena we perceive, and even the simple acknowledgement that humans are evolved animals and thus subject to the same inferences any other evolved species would be are grounding new insights into the phenomenon of civilization.

Given the present status of the social sciences, it is fair to contend that nearly every field is like a blind man, saying the part they are acquainted with means one thing, without realizing they are grasping only part of the whole. Economics, sociology, psychology, anthropology, even medicine are stuck in what is the equivalent of a geocentric model—but at least the geocentric models back in the day could provide accurate predictions, while the present standard model implicit of every discipline of the social sciences is continuously, relentlessly, ceaselessly, perpetually, unendingly falsified.

But of course it must be pointed out that the social sciences are not, in their present form, per se about science in the first place. Their primary purpose is to extol the praises of Progress and bash conservatives over the head with Science! because science is the lingua franca of knowledge, and you’re ignorant if you’d doubt what you read in the New York Times.

You Are a Bipedal Ape

Any explanation for human society that does not ultimately cohere to the kind of explanation you would find acceptable for explaining the conditions of the animal kingdom should be rejected out of hand, since it must necessarily void the most important and salient fact of the human condition; you are a bipedal ape. You know a few neat tricks which evolved into the human species through a hundred millennia of intraspecies conflict, and it is only several thousand generations which separate you from the hominid apes which are your ancestor. Nothing you do is ultimately qualitatively distinct from what you see animals doing. Instead of beating your chest like a gorilla would, you might engage in a loquacious disputation in order to assert your dominance over another trying to encroach on your territory, but the latter is different only by degree from the former.

The one neat trick that distinguishes humans from the rest of the animal kingdom is our unparalleled ability to cooperate. Our powers of cooperation were generated through a process of evolutionary means-end reversal; initially, all species were individual and solitary, but given the benefits of cooperation in game theoretic scenarios and the abundance of games that one plays in the animal kingdom, it was inevitable that some species would hack that ability to cooperate. The initial cooperation of our primordial ancestors eventually evolved to better capable forms of cooperation through the pressure placed on cooperation through intraspecies competition; after all, what could place greater pressures on the evolution of your species than that species which may best dominate the environment? This led to the development of language in order to better propagate social forms of knowing, enabling cooperation better directed to acknowledged and understood problems of material scarcity.

The human forms of behavior can and should be understood through the exact same analysis that one would make of other animal species. That simple shift in our ability and willingness to cooperate with others of our species opens many forms of environmental domination and subjugation not otherwise open to other species, but this is yet a consequence of our animal form and must be understood in relation to that. Call it misanthropology if you will: the study of humans as animals. The utility of such a perspective is no deviation from explanation by getting sidetracked on what you think should happen, and a mere observation of what does actually take place, the view you would take were you observing a struggling pack of wolves. You are permitted the ability to dispassionately point to those social forms would allow the production of solutions to the implicit collective action problems faced by the wolves. We are no different in this way, and collective action problems are the greatest problems humans face; mastering the control of matter is an easy feat next to the mastery of society.

The way which the individual learns is no different from that of pigeons and rats. We learn language, how to interact with people, how to do math, how to drive a car, and so on simply through positive and negative reinforcement, and tend to adapt reflexive behaviors which define our personality subconsciously as we find the strategies entailed by these reflexive behaviors are helpful for getting us what we want. Though we may have the ability to oversee our own learning process and so define for ourselves our conditions of learning, permitting second-order reinforcement of our learning is just another hack of our ability to cooperate. We learn instinctively, the majority of the channels of information which we receive are themselves social; what allows an individual of a species to become so radically dependent on his society for survival? Group selection entails that the success of an individual’s genes in being propagated is strongly tied to the success of that group, which entails ever greater orientation of the individual towards the group until the majority of the information he needs to concern himself with is social in content, and thus by necessity will be delivered through social means, which places the individual’s ability to verify several steps beyond him. Even the individual is dependent upon his society for knowledge, and the soundness with which that society distributes information is crucial to the soundness of that individual’s knowledge. Given knowledge is a prerequisite of more knowledge, it is only through the application of knowledge that structures of knowledge production may be checked and verified for their actual effectiveness in producing knowledge, leaving the individual in a knotty problem when he must determine which sources of knowledge really have a ‘’best fit’’ with the world. Indeed, the problem is so far beyond the individual that there is essentially nothing the individual can do to readily grasp the problems of society, which is why while we may have developed a rather keen sense of social sense for interaction at the interpersonal level, this knowledge of society at the interpersonal level does not scale; neither our language nor our forms knowing at the interpersonal level are apt to discussing society in the abstract, i.e. in terms of social causality which occur outside our direct acquaintance, and perceiving how the nature of individual relationships coheres within a system of relationships and the conditions under which they are formed.

Realizing you are a bipedal ape is key to understanding how little you really understand of how society works. What do you know that you do not know either through direct experience or through someone else telling you something? How well are you able to verify anything you are told? Even our standard of verifying our own knowledge tends to be heavily reliant on social conditions, e.g. able to be communicated to another. Knowledge is social, is constructed by society, and inasmuch as the individual is unable to check the mechanisms of social knowledge formation he is unable to check the content of his own knowledge for soundness and consistency.

Follow

Get every new post delivered to your Inbox.

Join 1,531 other followers